Development and Internal Validation of a Predictive Model for Operative Management in Blunt Abdominal Trauma Using Admission Physiological and Biochemical Parameters

Abstract

Background: Early identification of patients requiring operative management (OM) after blunt abdominal trauma is critical, yet initial physiological signs may be nonspecific. We sought to develop and internally validate an admission-based prediction model for early emergency department (ED) triage, prior to computed tomography (CT), using routinely available physiological and biochemical parameters. Methods: We conducted a retrospective observational study including adult patients with blunt abdominal trauma who underwent FAST and lactate testing at admission. OM was defined as any abdominal surgical intervention within 24 h to control hemorrhage or repair injury. A multivariable logistic regression model incorporating lactate, heart rate, leukocyte count, and FAST positivity was developed using complete-case data. Lactate diagnostic accuracy was assessed using ROC analysis. Internal validation was performed with 1000 bootstrap resamples. Results: In 81 patients with lactate results, lactate showed good discrimination for OM (AUC 0.815). At ≥3.5 mmol/L, sensitivity was 0.737 (95% CI 0.569–0.866), specificity 0.744 (0.588–0.865), LR+ 2.88, and LR− 0.35. The final logistic model demonstrated an apparent AUC of 0.904 and an optimism-corrected AUC of 0.882. The full model equation and coefficients are provided for reproducibility. Conclusions: Admission lactate, combined with FAST and physiologic measurements, provides useful early-triage information before CT and warrants external validation in larger cohorts.